22 BOTANICAL GAZETTE [JANUARY 
the impossibility of cutting exactly to the required depth, small 
parts of the cortex were probably left in many of the roots. 
Putting together the results of the last two columns, we see that 
more than 50 per cent. of the roots curved in some degree away 
from the brand. Hence the cutting did not inhibit the curvature 
as might be objected, but kept it below the cut in the large 
majority of cases. 
TABLE V. 
AXIAL CYLINDER OF ROOTS CUT OFF BY STABBING. 
— eum ro ay = Site tostab from stab 
1896 
16 2 
17 16 3 
ace 7 5 
19 II 8 
20 i1r2 fo) 10 4 
21 10 6 
21 9 5 10 8 
21 9 6 
nk 7 5 13 12 
24 9 7 
24 9 9 9 7 
25 9 5 
27 9 
Dee. 9 9 a 7 
7 10 4 
9 10 7 
9 10 3 
15 20 10 
: 20 14 
9 
March 13 9 6 
I I Il 
16 : 16 
30 14 66 45 208 125 
46.6 % 68.1 & 60.0 % 
Table V is instructive as evidence on the question of the 
transmission of the stimulus. All the roots had the axial cylin- 
der cut off as completely as possible, in such a way as to leave 
as much of the cortex as possible. The cutting was done with 
a long slender pointed scalpel, so narrow that it could be pushed 
through the root, cutting off the axial cylinder, but leaving most 
